Develop a tool that allows for inputs of various datasets, organizes the data with regards to specific behavioral parameters, and show interesting relationships between the data in intuitive visualizations.
Personal Health Risk Identifier
Develop an application/web-based tool that visualizes individual health risks by integrating personal data into pre-aggregated government data. The user inputs medical history / personal health metrics and the app returns cleanly visualized information about correlations of their behavior to population trends. The visualized information returned could be any or all of the following:
- your probability of certain health outcomes (diabetes, stroke, heart attack)
- the degree to which each parameter affects the probability of each outcome
- where you fall statistically compared to
- the US population, your state’s population, your community’s population (where available)
- others having the same age/race/gender/income/education
Visualizing Population Health Risks
Summarize relationships between various areas of data in compelling and meaningful ways – i.e. generate dynamic representations similar to the annual surveillance summaries CDC has published (example), though much more concise and aimed at the public. Conceivably, such reports should also be able to be auto-generated JIT based on some number of input parameters, so that the user could create a new report on an issue and population of interest to them.
Social Cost of Behavior
Compare relative treatment costs of health outcomes (e.g. lung cancer) from certain behaviors (e.g. smoking) in geographic regions with similar behavioral risk factor profiles.
Three projects from NHANES
We are looking for participants with advanced programming experience and data visualization skills (preferably in RStudio, Shiny, Tableau Desktop, or more advanced) to join us in visualization projects at NHANES in any of the following three areas. We will have a team of mentors available to help assist you in this project, and you will be invited to call into weekly meetings to discuss the status of the data and hear the latest information on how we are constructing data bases, selecting topics of public health interests, and build stories/themes to help better understanding of nation’s health.
We are greatly interested in those candidates who would be able to commit the full 3 months to create compelling products.
If interested, please contact Dr. Chong (email@example.com) for next steps and project roles.
Each of the following areas of evaluation would serve as a “project” submission, and each will be evaluated separately for prizes at the time of submission on October 28th. The three areas we are currently focusing on include:
- Visualizing relationships between height/weight and BMI (focusing on overweight and obese data)
- NHANES data can be used to visualize measured height, weight, BMI, waist circumferences, etc, by age, gender, race and education groups
- BRFFS can be used to show overweight and obesity trend by geographic location (states)
- Visualizing risk factors for cardiovascular diseases: cholesterol, hypertension, obesity, and diabetes, and comorbidity between these conditions (NHANES) as aggregated by geographic distributions (states) from BRFSS, and death rates with these conditions as multiple causes of deaths (vital statistics).
- Visualizing data from US Health Reports on major medical conditions using parameters that cross between the BRFSS and NHANES data
We think these visualizations look pretty cool.